Factors Affecting Delivery Location in Indonesia

Transkripsi

1 Factors Affecting Delivery Location in Indonesia

2 For information: MCHIP 1776 Massachusetts Avenue, NW Suite 300 Washington, D.C , USA Tel.: The Maternal and Child Health Integrated Program (MCHIP) is the USAID Bureau for Global Health s flagship maternal, neonatal and child health (MNCH) program. MCHIP supports programming in maternal, newborn and child health, immunization, family planning, malaria, nutrition and HIV/AIDS, and strongly encourages opportunities for integration. Cross-cutting technical areas include water, sanitation, hygiene, urban health and health systems strengthening. Visit to learn more. This study and report were made possible by the generous support of the American people through the United States Agency for International Development (USAID), under the terms of the Leader with Associates Cooperative Agreement GHS-A The contents are the responsibility of the Maternal and Child Health Integrated Program (MCHIP) and do not necessarily reflect the views of USAID or the United States Government. Prepared by: Sara Berthe Department of Maternal and Child Health Gillings School of Global Public Health University of North Carolina December 2011

3 TABLE OF CONTENTS ABBREVIATIONS AND ACRONYMS... ii INTRODUCTION... 1 LITERATURE REVIEW/EVIDENCE REVIEW... 2 METHODS... 4 Location... 4 Population... 4 Instrument Design... 5 Measures/Analytic Techniques... 6 RESULTS/FINDINGS... 8 Background Characteristics... 8 Delivery Location... 8 Reasons for Delivery Location Choice Cost of Delivery Services Physical Access Barriers Influencing Factors for Facility-Based Deliveries Disrespect and Abuse DISCUSSION/RECOMMENDATIONS Limitations and Strengths Next Steps REFERENCES...18 APPENDIX I: SURVEY INSTRUMENTS, ENGLISH...21 APPENDIX II: SURVEY INSTRUMENTS, BAHASA INDONESIA...41 Factors Affecting Delivery Location in Indonesia i

4 ABBREVIATIONS AND ACRONYMS AMDD ANC DHO IDHS MCH MCHIP MDG MMR MNCH SBA TBA USAID Averting Maternal Death and Disability Antenatal care District Health Officer Indonesia Demographic and Health Survey Maternal and child health Maternal and Child Health Integrated Program Millennium Development Goal Maternal morality ratio Maternal, neonatal and child health Skilled birth attendant Traditional birth attendant United States Agency for International Development ii Factors Affecting Delivery Location in Indonesia

5 INTRODUCTION As the 2015 deadline for the Millennium Development Goals (MDG) approaches, Indonesia still suffers from a high maternal mortality ratio (MMR) of 228 per 100,000 live births. 1 The fifth MDG goal, which seeks a reduction in maternal mortality by 75% between 1990 and 2015, looks to Indonesia to reduce its MMR from an initial 421 per 100,000, as cited in the 1991 Indonesia Demographic and Health Survey (IDHS), to 100 per 100, While the government has previously implemented several population-level programs that have reduced the MMR, Indonesia s large and diverse population, wealth disparity and urban/rural divide continue to contribute to maternal deaths. Additionally, in the wake of the 1998 and 2004 government decentralization efforts, decisions were not automatically handed down from the national government in Jakarta, thereby allowing local governments more say and choice in their investments. 3 Local funding is often first spent on visible and tangible infrastructure improvement, with remaining money allocated to services such as education and health. This situation has increased the health disparities between regions, as certain regions may put more emphasis on maternal health than others. MDG 5 seeks to improve maternal health, both by reducing by three-quarters the MMR and by increasing universal access to reproductive health care. In Indonesia, women die every hour from complications during pregnancy and delivery a problem exacerbated by the fact that 53% of deliveries take place at home and 24% of deliveries are assisted by a traditional birth attendant (TBA) with no formal training. 1,4 Women s decisions on where to deliver their children are based on a variety of factors including cost, distance to facility and the type of provider available to assist them. The goal of this study was to increase our understanding of the reasons women choose either to stay home or deliver in a facility, and to determine the breakdown of facilities used in the public sector compared to the private sector. Overall, this study will inform programmatic decision-making in an effort to reduce barriers to access and increase the number of women choosing to deliver at a facility. Factors Affecting Delivery Location in Indonesia 1

6 LITERATURE REVIEW/EVIDENCE REVIEW In developing countries, especially those in Southeast Asia and sub-saharan Africa, cultural and societal norms lead to high levels of home-based deliveries. Although exact figures vary depending on the country, and on the regions within the country, a study based on data from 48 DHS surveys spanning the globe found that more than half of deliveries took place at home. 5 Home deliveries are also most common among the poor. 5 The link between mother s education and skilled care has been studied extensively and we know that higher levels of education lead to more autonomy and decision-making power, a greater ability to acquire and understand knowledge, and improved attitudes toward health problems and facilities. 6,7,8 Mothers education is an important predictor for home vs. facility-based birth. 9 Studies in both Bangladesh and Nepal note that low education levels are associated with women delivering at home. 9,10 The research on delivery location in Nepal also raises multi-parity as an indicator of home delivery, stating that the more children a woman has, the more likely she is to continue delivering at home. 10 Additionally, many women consider their previous birthing experience as a proxy for future deliveries; therefore, if they delivered at home and had no complications, they perceive that all future births can safely occur at home. 11,12 Women in Indonesia and Timor Leste cited birth complications as the only reason they would travel to a facility. 11,12 Cost is a major barrier in women s access to delivery services. Some countries, including Indonesia, have created government insurance schemes that will pay for the delivery services for poor women. In January 2011, the Government of Indonesia created a new public assistance insurance scheme called jampersal, which allows all women, regardless of level of income, to deliver for free at a facility. 13 Unfortunately, the government has not yet uniformly rolled this program out to all of the districts. Jampersal, and other insurance schemes, aim to alleviate the burden of cost on the woman, and promote higher rates of facility-based and therefore skilled provider, deliveries. Despite these schemes, there are other costs associated with leaving one s home to deliver in a facility, including the cost of transportation, food and childcare 11,12. A study in West Java, Indonesia, showed that, even with delivery at a health facility available free of charge, women preferred the services of TBAs or were either confused or embarrassed about using the government insurance program and ultimately delivered at home In many developing countries, distance to a health facility is a major factor in the selection of delivery care services. 15,16,17 Because of poor infrastructure, it often takes women hours or even days to arrive at a health facility, which encourages women to stay at home to deliver their babies. The issues of distance and transportation often overlap because transportation is very unreliable and often women are on foot. Skilled birth attendants (SBA) noted that distance and transportation in Bangladesh make it challenging to reach women s homes in time for the delivery. 18 Accordingly, we can extrapolate to infer that the women in question would likely have difficulties reaching a health facility. The studies mentioned have strengths and limitations in their analysis of factors affecting delivery location. The study population is different in all studies, and some are more representative than others. The data from Timor Leste and Indonesia sought specifically unusual cases and focused on one province in each country that is known to rely heavily on TBAs. 11,12 Therefore, these results may not be generalizable to the entire population of these countries, nor to the greater Southeast Asia region. The sample size from Matlab, Bangladesh, was greater than 41,000 people, but because Matlab has been a research center since the 1980s, its residents may be influenced by its use as a study site and therefore may not be representative of the larger population. Finally, by changing the perspective and interviewing SBAs instead of women, another study in Bangladesh attempted to prove that home-based 2 Factors Affecting Delivery Location in Indonesia

7 deliveries are dangerous, due to lack of equipment and medicine, and inconvenient for the SBAs, and women therefore should travel to health facilities to deliver. 18 This study seeks to understand the reasons that women choose non-facility-based births instead of facility-based births. It adds several new components to previous studies, including questions about whether the quality of services and knowledge of maternal and child safety influence a woman s decision to choose a facility. Given the perceived growing trend in facility-based disrespect and abuse, women were also asked to report on disrespect, mistreatment and abuse that they experienced, witnessed or heard during their delivery. At present, no other research has been published that looks at women s perceptions of non-facility-based deliveries versus facility-based deliveries in the same timeframe and location as this research. Factors Affecting Delivery Location in Indonesia 3

8 METHODS The dependent variable of this study is the delivery location, which has five possible outcomes: home, puskesmas, private clinic, public hospital and private hospital (see definitions below). The mediator variables that directly affect women s choice of delivery location include cost of delivery and physical access barriers, such as distance to facility and availability of transportation. Furthermore, there are several independent variables including level of education, level of income, age, availability of insurance (either publicly provided by the government or privately obtained), and knowledge and attitudes about maternal and child health (MCH). In this model, age, level of education and level of income play a central role as the basis for other independent variables that affect women s choice. The findings from this study will inform programmatic decision-making for the Maternal and Child Health Integrated Program (MCHIP) and other Jhpiego programs including the relationship between influencing factors and the delivery location. MCHIP s goal is to analyze the influencing factors to determine what, if any, programs can be implemented to increase women s knowledge and reduce barriers to access. LOCATION This study was designed with the programmatic needs of MCHIP in mind, and the data were collected by the author, with assistance from a translator who was formally trained as a midwife and therefore had intimate knowledge of the subject matter. The survey was conducted in four districts in Indonesia where MCHIP currently works. By working in these districts, Jhipego had already established a relationship with the District Health Offices, and the mission for the project was pre-determined. The four districts spanned the islands of Sumatra and Java and included Minas, Karawang, Bojonegoro and Serang. In Minas, Karawang and Serang, we surveyed women in six villages, and in Bojonegoro we surveyed women in five villages. Within each district, we coordinated with the District Health Officer (DHO) to choose the villages and facilities where the interviews would be conducted. Villages were purposively selected in collaboration with the DHO and village midwives, based on at least one of the following criteria: availability and interest of the village midwife in coordinating a gathering of woman for the survey; villages facilitating mothers centering (support) groups; and villages with an antenatal care (ANC) clinic. POPULATION The participants in this study included three predetermined groups of women: pregnant women, postpartum women who delivered in a facility and postpartum women who delivered at home or in a facility not covered in this study. Postpartum was defined as no more than six months after delivery. For our purposes, facility was defined as one of the following types of health care posts: puskesmas, or community health center established and funded by the government and including at least one midwife, nurse and doctor; private clinic, which included private midwifery practices or private clinics that employed at least one midwife and one doctor; public hospital, established and paid for with government funding; and private hospital, established and paid for by private donors. Non-facility was defined as: home or pustu (sub-health center), poskesdes (village health post) or polindes (village maternity post) all of which generally have little to no equipment and often a midwife or nurse as the only health care provider regularly available. These distinctions were defined with the help of MCHIP/Jhpiego staff, and those facilities that are included in the non-facility category were deemed not to have sufficient equipment for deliveries, or for emergency obstetric care in the case of complications. In our third district, Bojonegoro, the district-level government invested a lot of money in the polindes, which were often better equipped than some of the puskesmas. Despite this disparity, and due 4 Factors Affecting Delivery Location in Indonesia

9 to the country-wide precedent of polindes usually having fewer resources than puskesmas, we continued to include them as non-facilities. In each district, we sought to interview women per day, with the goal of interviewing 100 women from each group. After the DHO selected the villages, the village midwives were charged with finding approximately 15 pregnant and postpartum women who fit the parameters of our study. All women identified who met the study criteria for pregnancy/postpartum status were eligible to be interviewed, regardless of age, parity and the like. The sample was a convenience sample, as participants were selected at the convenience of the village midwife, and is therefore not a statistically representative sample of the population. We were in close communication with the DHO and village midwives one week before our arrival in each district, and therefore the village midwives had time to consult with pregnant and postpartum women and invite them come to the meeting place (often a health post) at a specified time and place. While the village midwife attempted to find equal numbers of pregnant women, non-facility delivery postpartum women and facility-delivery postpartum women, she was not always able to do so. Our final number of participants was 300 and the breakdown included 93 pregnant women, 91 postpartum non-facility deliveries and 116 postpartum facility deliveries. We were unable to meet the goal of 100 women per group because of women s unavailability or their postpartum status falling outside of the sixmonth timeframe. All participants were all provided with snacks while waiting for the interview to begin and then received a wellness kit, including a toothbrush, toothpaste, soap and washcloth, when the interview was complete. No monetary compensation was provided. An overview of the study, with a request for consent, was read to each participant, and each woman signed the informed consent document allowing us to begin the interview. INSTRUMENT DESIGN The instrument for this study was a paper-and-pencil style questionnaire that the study team created. The document was originally developed in English and translated into Bahasa Indonesia. When interviewing the participants, the interviewer gave each woman an identification number, and only information relating to the district and village where the interview took place was collected. Each question and, where applicable, list of responses was read to the participant to avoid any literacy-related problems and to ensure that questions were read the same way each time. The questions are a combination of previously validated questions from the IDHS, questions about disrespect and abuse from the Averting Maternal Death and Disability (AMDD) Program questionnaire used during facility exit interviews in Tanzania and questions that we ourselves created. After conducting the literature review, and reviewing other surveys that sought data on quality of care and the reasons that women choose where to seek medical care, we created the remaining questions. We sought the answers to questions around birth and delivery planning, including delivery location, delivery assistant (TBA, midwife, ob/gyn, etc.), cost and distance to facility. We asked about presence of any birth complications and if participants were referred to higher-level medical establishments for these complications. The primary research question was to identify the reasons that women chose to deliver at the location of their choice. We asked all postpartum women where they delivered and followed up by asking, Why did you choose to deliver at that location? We then asked all postpartum women who did not deliver at a facility, and all pregnant women, Which of the following would encourage you to go to a facility to deliver? This list of possible responses originally had 16 options, which were read to all respondents, including an other category, from which women could choose as many options as were applicable to them or give a reason that was not included. When the other responses were listed, the number of choices increased to 24. These response options were consolidated into eight categories for analysis, as seen in Table 1. Factors Affecting Delivery Location in Indonesia 5

10 Table 1: Consolidation of Reasons for Choosing Delivery Location Cost Insurance Distance Transport available ORIGINAL QUESTION OPTIONS Cleanliness Good supply of drugs and equipment Type of health provider Experience of health provider Attitude of health provider Relationship with health provider Being treated with respect Midwife unavailable Safety for mother and child Afraid Health worker recommended Referred by another facility Recommendation from family Facility where usually go Privacy Comfort Childcare Weather Didn t make it in time Nothing ANALYSIS CATEGORIES Cost Physical Access Quality of Services Knowledge Recommendations Convenience Other Nothing For the majority of the 300 participants, this question was difficult to understand. For the first two districts that we visited, the same response options were provided for all women. For the last two districts, we amended the responses to facilitate understanding, and created separate responses for those who had delivered or were choosing to deliver in a facility and for those who had not chosen or were choosing not to deliver in a facility. For example, instead of asking if distance was the reason a woman either chose a facility or did not, we asked women who chose a facility if the facility was close to your home, and women who didn t choose a facility if the facility was far from your home. With the former wording, women had a difficult time understanding this question and the translator would have to ask it multiple times in order to receive a response. With the wording amended, women were more easily able to understand and answer this question. MEASURES/ANALYTIC TECHNIQUES The overall dependent variable in this study is delivery location. The independent variables that we will explore as affecting this variable are age and education level, with cost of delivery and distance to facility as mediating factors. We will further explore how the questions around the perception of influencing factors affected the choice of delivery location. All quantitative analyses were performed using Stata SE 12. Data were primarily analyzed using descriptive statistics, specifically cross-tabulations. With a relatively small total sample size, the sample size was very small for some subgroups, such as older women, women with a secondary-level education and those making more than 2 million INR per year. To reduce high sampling errors associated with these small sample sizes, certain categories were collapsed. 6 Factors Affecting Delivery Location in Indonesia

11 Age ranges were collapsed into 15 24, and 40 49, education collapsed to some elementary, some junior high and senior high/secondary, and income collapsed to low, mid and high. This study was exempt from Institutional Review Board (IRB) approval on the basis that it was considered to be a programmatic study informing program implementation rather than contributing to generalizable knowledge. Factors Affecting Delivery Location in Indonesia 7

12 RESULTS/FINDINGS BACKGROUND CHARACTERISTICS Descriptive statistics for background characteristics are presented in Table 2. The mean age of respondents was 28 and the majority of the respondents fell in the age range of Most women had either some primary or some junior high school education, but relatively few had any senior high school or secondary education experience. In this predominantly Muslim society, all 300 respondents were married women. Data on religion and ethnicity are not shown; however, 278 (92%) of the women self-identified as Muslim, while the remaining 22 (7%) selfidentified as Protestant, Catholic or Confucian. Additionally, in this ethnically heterogeneous country, 92% of the respondents were Javanese, Batak or Sundanese, and the remaining 7% were Malay, Minang, Nias and Aceh (data not shown). Table 2: Background Characteristics by District MINAS KARAWANG BOJONEGORO SERANG TOTAL AGE AND EDUCATION % n % n % n % n n Age Don't Know Total Education Primary Jr. High Sr. High/Secondary Total DELIVERY LOCATION All postpartum respondents reported the location of their most recent delivery, while pregnant women reported the planned location of their upcoming delivery. The results presented in Table 3 contrast the actual and planned delivery locations of all 300 respondents. The postpartum data combine facility and non-facility births. Of the postpartum women, 24% (n=49) chose a private clinic as their planned location, followed by 19% (n=40) of women choosing puskesmas. The 93 pregnant respondents were almost evenly divided in their choice of planned delivery location, with 36% (n=34) responding that they would deliver in a private clinic and 32% (n=30) reporting that they would deliver at home. 8 Factors Affecting Delivery Location in Indonesia

13 Table 3: Planned and Actual Delivery Locations PREGNANT (PLANNED) POSTPARTUM (ACTUAL) PLACE OF DELIVERY % n % n Home Puskesmas Private Clinic Hospital Private Hospital Other Total Table 3 identifies delivery location by the respondents background characteristics, providing more detail by district, age, education level and income level. Breaking down these background characteristics by district shows the disparities among the districts. Overall, home deliveries accounted for almost 34% of all postpartum responses, followed by private clinics at 23.6%. For home deliveries, Serang and Karawang had the highest incidences, while Bojonegoro had the lowest incidence of home deliveries, but the highest incidence of other non-facility deliveries, most of which occurred at the polindes. Puskesmas and private clinic deliveries fell around the same range across all districts. The highest percentages of home deliveries occurred for women between years old and those with some elementary education. Hospital deliveries were rare across districts, and often occurred only for those women with birth complications. In terms of income, the highest percentage of low-income women delivered at home, while the highest percentage of high-income women delivered at a private clinic. While the total sample population in this table is small, it does speak to the differences in background characteristics among districts and may help to explain the reasons that influence women s decisions regarding their delivery location (see Table 4). Factors Affecting Delivery Location in Indonesia 9

14 Table 4: Actual Place of Delivery by Background Characteristics Postpartum Women HOME PUSKESMAS PRIVATE CLINIC PUBLIC HOSPITAL PRIVATE HOSPITAL OTHER TOTAL BACKGROUND CHARACTERISTICS % n % n % n % n % n % n % n District Minas Karawang Bojonegoro Serang Age Education Elementary Junior High School Senior High/Secondary Income Level Low Middle High Total Factors Affecting Delivery Location in Indonesia

15 REASONS FOR DELIVERY LOCATION CHOICE Women in this study chose their delivery location for a variety of reasons, including cost, level of quality, physical access, knowledge, recommendations and convenience. Women chose as many of these categories as were applicable to them and Figure 1, below, shows the distribution of these responses. For facility deliveries, most women responded that quality of services (73%), physical access (44%) and their own knowledge (42%) were the top reasons for choosing a facility. These women knew and understood the quality of services provided at facilities, including experience of health care providers, supplies of drugs and equipment, and cleanliness. Physical access, in terms of distance to facility location and availability of transportation, was not a barrier for them. Finally, their own knowledge that facility deliveries were safer for both mother and child prompted their decision. Figure 1. Reasons for Delivery Location Percentage of Women Facility Non-Facility The postpartum women with non-facility delivery cited cost (46%), convenience (45%) and physical access (43%) as their reasons for choosing a non-facility delivery. Because most of the non-facility deliveries in this survey occurred at home, where childbirth is presumably less expensive, it is understandable that cost was a barrier in the women s ability to seek care at a facility. The convenience category, as mentioned above in Figure 1, is composed of: usual facility, comfort and childcare. The women who chose convenience stated that they previously delivered at home with no complications and therefore chose their home again, wanted their families surrounding them at the time of birth and had other responsibilities, such as childcare, that they could not abandon. Physical access also was a barrier for these women in seeking care at facilities, in that they lived too far from a facility or did not have adequate transportation available. These data are important and tell us two things: that women who are going to facilities understand that quality of care and maternal and child health and safety are important, and that cost and physical access barriers are preventing other women from seeking the same services. Factors Affecting Delivery Location in Indonesia 11

16 COST OF DELIVERY SERVICES Postpartum women in our survey were asked to recall the cost of their deliveries. Note that this total includes the cost of paying the provider as well as any medications needed during the delivery, but does not include auxiliary costs, such as the cost of transportation and food while at their delivery location. Cost of delivery is broken down into the following categories: Free, 1 250,000 INR, 250, ,000 INR, 500,000 1 million INR, and more than 1 million INR. At the time of this survey, those figures translated to less than $21, $21 $42, $42 $83 and greater than $83, respectively. Figures 2 and 3 below show the breakdown of costs for non-facility and facility deliveries, respectively. While no-cost deliveries were somewhat common in both populations, these were often the result of women using either government or private insurance, and were not the result of not paying for the delivery. In some cases, women citing extreme poverty who delivered at home with a TBA paid in material goods such as fabric, livestock or rice. Of the 207 postpartum women, there were only three such cases. The cost of delivery for non-facility births also varied depending on the provider. Twenty non-facility deliveries occurred with a TBA and 44 with a midwife (data not shown). It is interesting to note that the cost to deliver with a midwife was the same whether the delivery was performed at the pregnant woman s home or at the midwife s private facility. This was the case in all districts. Across all non-facility births attended to by either a TBA or a midwife, no women delivering with a TBA had no-cost deliveries, while eight women delivering with midwives did. In four cases, women delivering with TBAs paid the same cost as women who delivered with a midwife. If perceived cost is the greatest barrier to facility-based care, increased education for pregnant women should convey that costs for facility-based delivery with a midwife can be less expensive than, or equal to, the cost of delivering at home. PHYSICAL ACCESS BARRIERS In many developing countries, distance to facilities and access to transportation are two barriers that limit women from accessing health care services. For non-facility deliveries, women cited physical access barriers as the second most common reason for not seeking facility-based care. All participants in this study were asked about their proximity to the closest facility, the availability of transportation and the cost of transportation. The distance question was asked in terms of both kilometers and time, and although time is a more subjective measure, more women were able to accurately answer the question about time than that about distance. Figure 2: Non-Facility Cost of Delivery Figure 3: Facility Cost of Delivery Free 1-250K K 500K-1M Free 1-250K K 500K -1M >1M Don't Know 12 Factors Affecting Delivery Location in Indonesia

17 Table 5 presents the data on physical access barriers. Of the 91 women delivering in a nonfacility, 52% of them said that the closest facility was less than 15 minutes from their home and 13% said it was between 15 and 30 minutes. Note that this is measured in time only and does not take into account the mode of transit, such as walking versus using a motorcycle. Only 4% lived greater than 30 minutes from a facility, but 30% were unable to answer the question. Women were also asked if they or their family had access to a motorcycle or car, and 97% said that they did. In terms of the cost of transportation, 67% of these women said that the cost of traveling to the closest health facility was less than 10,000 IDR (approximately $1.15), while 21% said that it would cost between 10,00 50,000 INR ($1.15 $5.65). Table 5: Physical Access Non-Facility Deliveries DISTANCE TO CLOSEST FACILITY (MINUTES) % n COST OF TRANSPORT TO CLOSEST FACILITY (INR) % n DO YOU HAVE ACCESS TO A MOTORCYCLE OR CAR? % n < Free Yes <10, No ,000 50, , , < >100, Don t Know INFLUENCING FACTORS FOR FACILITY-BASED DELIVERIES This study sought to identify the factors that would encourage women to choose facility-based care. We asked all pregnant women who were planning on a non-facility delivery and all postpartum women who had recently had a non-facility delivery, Which of the following would encourage you to go to a facility for delivery? using the same response categories mentioned above. Women chose as many options as were applicable to them. Table 6 describes the factors that pregnant and postpartum women cited would influence their decision to seek facility-based care for deliveries. Cost, quality of services and physical access were the top reasons cited by both groups. This essentially means that if cost were reduced, quality of services increased and physical access barriers removed, these women would be more likely to choose facility-based care for their deliveries. Table 6: Influencing Factors That Would Encourage Non-Facility Deliveries to Choose a Facility, % FACTORS PREGNANT POSTPARTUM Cost Quality Physical Knowledge Recommendation Convenience Other 0 0 Nothing Factors Affecting Delivery Location in Indonesia 13

18 Increased knowledge of maternal and child safety is an important influential factor for postpartum women, but negligible for pregnant women. Of interest as well is the percentage of women who cite that nothing would influence their decisions to deliver in a facility. These percentages are relatively low, but speak to the fact that no matter the intervention, some women will continue to deliver at home. DISRESPECT AND ABUSE A current trend in research looks at the incidence of disrespect and abuse at facilities as it relates to behaviors in seeking facility-based care. We asked postpartum women, regardless of delivery location, if they were treated in any way that made them feel disrespected during their delivery. Only 3.4% of these women felt disrespected (data not shown). For all postpartum and pregnant women, we also asked if they experienced, witnessed or heard about any of the following abuses occurring anytime at a facility: being shouted at or scolded; being hit, beaten or slapped; treatment threatened to be withheld because a woman couldn t pay; threatening for any other reason; abandonment when care was needed; and not being able to leave the hospital with the baby because of failure to pay. Women most often said that they had heard about one of these abuses happening, specifically shouting or scolding and abandonment during care, but instances of experiencing or witnessing these abuses were rare or nonexistent (see Table 7). Therefore, there is little reason to assume that disrespect or abuse at facilities relates to decision by these women to seek facility-based deliveries. Table 7: Report of Disrespect and Abuse, by District MINAS KARAWANG BOJONEGORO SERANG TYPES OF DISRESPECT/ABUSE Disrespect % n % n % n % n Yes No Total Shouting or Scolding Experienced Witnessed Heard About None of the Above Total Hitting, Slapping, Pinching Experienced Witnessed Heard About None of the Above Total Factors Affecting Delivery Location in Indonesia

19 MINAS KARAWANG BOJONEGORO SERANG TYPES OF DISRESPECT/ABUSE Threaten to Withhold Treatment % n % n % n % n Experienced Witnessed Heard About None of the Above Total Threaten for Other Reasons Experienced Witnessed Heard About None of the Above Total Abandonment Experienced Witnessed Heard About None of the Above Total Unable to Leave Hospital Due to Failure to Pay Experienced Witnessed Heard About None of the Above Total Factors Affecting Delivery Location in Indonesia 15

20 DISCUSSION/RECOMMENDATIONS Many factors play into a woman s decision on where to seek delivery services during her pregnancy. Previous literature has shown a strong link among age, level of education, cost of delivery, and delivery location. This study supports the associations between level of education and delivery location and cost of delivery and delivery location, but does not support the association between age and delivery location. In this study, we found that women who are seeking facility-based care are doing so because they seek a higher quality of services and know that they and their child will be safer. It is important to note that even though cost was not one of the main factors influencing their decision, it was still important. Many of the facility-based deliveries were paid for either with jampersal or another insurance covering some or all of the cost. Had insurance not been available to these women, it is possible that they would have chosen non-facility deliveries. For women not seeking facility-based deliveries, cost and physical access were the two main barriers. Women s perceptions of the cost of delivery and the distance to their closest facility persuaded them to deliver either at home or at a less-equipped facility. The government needs to scale up their rollout of jampersal and ensure that all districts in Indonesia are aware of the program. During this study, we found that jampersal was available in only two of the four districts, and was being improperly used for home deliveries in one of those districts. A media campaign would further inform women and their communities about this service. Without this free delivery service available, and because cost is a main factor in choosing delivery location, women will likely continue to deliver at home, and these homebased births will continue to contribute to maternal and infant mortality in Indonesia. With distance as a perceived barrier, but with the data showing that most facilities are within 30 minutes of a woman s home, the government should sensitize the population so that each household knows the closest facility available to them. Additionally, although transportation cost itself was not identified as a barrier, the government or organizations within these communities should consider a low-cost or free transportation system for pregnant women. Because a large percentage of women cited an improvement in quality of services as an influential factor that would encourage them to seek facility-based deliveries, ensuring a minimum quality standard in all facilities is essential. In Minas, one village that was 20 kilometers from the closest puskesmas had established a community transportation system. Women who sought to deliver at the puskesmas were provided free transportation. With government or non-governmental assistance, similar programs can be implemented to give more women opportunities for facility-based delivery. Since 97% of women had access to some type of transportation, and travel time to facilities did not often exceed 30 minutes, the perception of physical access barriers appears to be exaggerated. Additional reasons must be compounding a woman s decision to seek a non-facility-based delivery. Education campaigns informing women of the dangers of home delivery already exist, but they should be scaled up and continued. They can include the pregnant women s educational classes currently being conducted; in addition, midwives and community lay health workers should be educated and trained so they can in turn educate clients as well. We also learned that although disrespect and abuse at facilities exist, they may be more hearsay than events that are experienced or witnessed. There is no indication that either of these factors influenced delivery location among the women in this survey. 16 Factors Affecting Delivery Location in Indonesia